RESEARCH ARTICLE

Mercury-Pollution Induction of IntracellularLipid Accumulation and LysosomalCompartment Amplification in the BenthicForaminifer Ammonia parkinsonianaFabrizio Frontalini1*, Davide Curzi2, Erica Cesarini2, Barbara Canonico2, FrancescoM. Giordano2, Rita De Matteis2, Joan M. Bernhard3, Nadia Pieretti1, Baohua Gu4, JeremyR. Eskelsen4, Aaron M. Jubb4, Linduo Zhao4, Eric M. Pierce4, Pietro Gobbi2,Stefano Papa2, Rodolfo Coccioni1

1 Department of Pure and Applied Sciences, Urbino University, Urbino, Italy, 2 Department of BiomolecularSciences, Urbino University, Urbino, Italy, 3 Geology and Geophysics Department, Woods HoleOceanographic Institution, Woods Hole, MA, United States of America, 4 Environmental Sciences Division,Oak Ridge National Laboratory, Oak Ridge, TN, United States of America

* fabrizio.frontalini@uniurb.it

AbstractHeavy metals such as mercury (Hg) pose a significant health hazard through bioaccumula-

tion and biomagnification. By penetrating cell membranes, heavy metal ions may lead to

pathological conditions. Here we examined the responses of Ammonia parkinsoniana, abenthic foraminiferan, to different concentrations of Hg in the artificial sea water. Confocal

images of untreated and treated specimens using fluorescent probes (Nile Red and Acri-

dine Orange) provided an opportunity for visualizing the intracellular lipid accumulation and

acidic compartment regulation. With increased Hg over time, we observed an increased

number of lipid droplets, which may have acted as a detoxifying organelle where Hg is

sequestered and biologically inactivated. Further, Hg seems to promote the proliferation of

lysosomes both in terms of number and dimension that, at the highest level of Hg, resulted

in cell death. We report, for the first time, the presence of Hg within the foraminiferal cell: at

the basal part of pores, in the organic linings of the foramen/septa, and as cytoplasmic

accumulations.

IntroductionForaminifera are single–celled organisms (protists) with a shell (i.e., test) that may be organic,agglutinated, composed of secreted calcium carbonate or, in rare cases, silica [1]. They are con-sidered one of the most diverse group of shelled microorganisms in modern oceans [2] andplay a significant role in global biogeochemical cycles of inorganic and organic compounds [3].Traditionally, the study of foraminifera has been the domain of paleontologists that have usedthem for petroleum exploration surveys, and palaeoclimatological and palaeoecological

PLOSONE | DOI:10.1371/journal.pone.0162401 September 7, 2016 1 / 14

a11111

OPEN ACCESS

Citation: Frontalini F, Curzi D, Cesarini E, CanonicoB, Giordano FM, De Matteis R, et al. (2016) Mercury-Pollution Induction of Intracellular Lipid Accumulationand Lysosomal Compartment Amplification in theBenthic Foraminifer Ammonia parkinsoniana. PLoSONE 11(9): e0162401. doi:10.1371/journal.pone.0162401

Editor: Yi Hu, Chinese Academy of Sciences, CHINA

Received: April 19, 2016

Accepted: July 26, 2016

Published: September 7, 2016

Copyright: This is an open access article, free of allcopyright, and may be freely reproduced, distributed,transmitted, modified, built upon, or otherwise usedby anyone for any lawful purpose. The work is madeavailable under the Creative Commons CC0 publicdomain dedication.

Data Availability Statement: All relevant data arewithin the paper and its Supporting Information files.

Funding: The research for this paper was partiallymade possible by the financial support from the PRIN2010-2011 Ministero dell’Istruzione, dell’Università edella Ricerca (MIUR) (protocollo 2010RMTLYR) toRC. JMB acknowledges support from The Investmentin Science Fund at WHOI. BG, JRE, AJ, LZ, andEMP were supported in part by the Office ofBiological and Environmental Research, USDepartment of Energy (DOE) as part of the MercuryScience Focus Area at Oak Ridge National

reconstructions. Benthic foraminifera have also been used as bioindicators of pollution in avariety of marine and transitional marine environments (i.e., [4–5]). They respond to adverseecological conditions including pollution by changing assemblage compositions and parame-ters (diversity and density) as well as cellular ultrastructure (reviewed in [6–7]). Despite majoradvances on foraminiferal ecology and biology that have been achieved over 50–60 years, theultrastructure-induced changes due to pollution [8–11] are not fully understood. An increaseof lipid droplets (LD) characterized by a more electron-dense core, proliferation of residualbodies, a thickening of the organic lining, mitochondrial degeneration and autophagosomeproliferation are among the main cytological alterations of benthic foraminifera in response topollutant exposure [9–11].

The proliferation of fibrillar and large lipidic vesicles, their possible exocytosis and anincrease in the number of residual bodies have been documented in two Ammonia species afterexposure to different copper (Cu) concentrations [10]. Interestingly, those authors reportedthe presence of larger lipid vesicles in all chambers of specimens exposed to higher Cu; theseresults were interpreted as a perturbation in the foraminiferal metabolic regulation due to pol-lution. The thickening of the organic lining might be ascribed to fibrillar material providedthrough fibrillar vesicles that discharge their content by exocytosis (i.e., [7,12]). It was sug-gested that the thickening of the organic lining might have a defense mechanism against pollut-ants [6–7]. A higher number of LD was noted in Ammonia parkinsoniana exposed to highconcentrations of lead (Pb); these LD were also characterized by an electron-dense core thatwas not, however, visible in the LD from control specimens [11]. Experiments on Rosalina leeirevealed that increasing concentrations of Hg promote the inhibition of reticulopodial activity,and more importantly, the development of test abnormalities as well as a reduction of growth[13–14]. Despite these observations, no ultrastructural studies have been performed on benthicforaminifera exposed to Hg. Mercury and most of its compounds are extremely toxic and areconsidered among the most harmful toxins for living organisms [15–16]. Mercury is known tobioaccumulate and biomagnify as methylmercury (CH3Hg+) in aquatic organisms [17].

The aim of this contribution was, therefore, to assess the variation in the distribution andamount of lipids in A. parkinsoniana when exposed to different concentrations of Hg and tolocalize the presence of this pollutant within the cytoplasm. This was achieved using transmis-sion electron (TEM), environmental scanning electron (ESEM), and scanning electron (SEM)microscopies coupled with energy dispersive X-ray spectrometry (EDS), and confocal laserscanning microscopy (CLSM) with fluorescent dyes.

Materials and Methods

Sediment sampling and laboratory mercury treatmentSediment samples were collected at 14-m water depth off the Monte Conero area (Italy, AdriaticSea). The collection site (43°33'54'' N, 13°39'52'' E) is in a coastal area located close to the terres-trial Regional Natural Park of Conero characterized by little human activity, with largely diversi-fied benthic foraminiferal assemblages and oligo-mesotrophic conditions [18]. No permits orapprovals were required to collect sediments at this site. No threatened or endangered specieswere involved. At the collection site, temperature, pH, salinity, Eh and dissolved oxygen of seawa-ter were measured in vertical profile using a multiparameter CTD (Conductivity, Temperatureand Depth) probe (Hydrolab, MiniSonde 4a). Sediment was sampled by multiple deployments,at the same station, of a Van Veen grab sampler that collects sediment over a surface area ofabout 400 cm2 and only the surface 2 cm were retained. On board, the sediment was thoroughlyhomogenized and subsequently sieved over a 500-μm screen with ambient seawater. The>500-μm fraction was discarded to remove potentially disturbing effects of bioturbators (i.e.,

Mercury-Pollution and Ammonia parkinsoniana

PLOSONE | DOI:10.1371/journal.pone.0162401 September 7, 2016 2 / 14

Laboratory, which is managed by UT-Battelle LLC forthe DOE under contract DE-AC05-00OR22725.

Competing Interests: The authors have declaredthat no competing interests exist.

macrofauna and large meiofauna). The<500-μm fraction, which bore foraminifera, was placedin an insulated box, covered with ambient seawater, and kept near ambient temperature untilarrival at our shore-based laboratory. The study did not involve any animal (animals, embryos ortissues) or human (human participants and/or tissue) participation.

Artificial Sea Water (ASW), which was prepared at in-situ temperature following the meth-ods of [19], was stored in the dark and oxygenated. Two Hg-ASWmixture concentrations plusone control (i.e., ASW without added Hg) were prepared. The inorganic salt of Hg, as mercurychloride (HgCl2;>99.5% pure; CAS Number 7487-94-7; Sigma-Aldrich), was used for theexperiments. The final pollutant concentrations for experimental media were obtained by add-ing appropriate volumes of stock solutions to ASW. High concentrations of Hg, namely 1 ppmand 100 ppm, were considered for the experiment to ensure that contaminant permeated thesediment and, in turn, incubated the benthic foraminiferal assemblages. Each tank (aquarium)(60 cm × 40 cm × 20 cm) was filled with 20L of Hg-ASWmixture. Mesocosms (15 cm × 8cm × 3 cm) containing 1 cm-thick sediment were placed inside each tank. Multichannelpumps were used to circulate and to oxygenate water through silicone rubber tubing anchoredbetween the tanks’ bottom and plastic grids. Tanks were placed in a controlled environmentwith air temperatures of 14–16°C that were uniformly maintained throughout the experiment.The dissolved oxygen (DO), salinity (S), conductivity, temperature (T), Oxidation ReductionPotential (ORP) and pH of the seawater were routinely monitored by a set of HQ40d (HachLange) portable multi-parameter probes [11]. Physico-chemical parameters (DO, S, T, pH andORP) of seawater were kept mostly constant throughout the experiments. The mean value ofsalinity registered during the experiment was 36.98‰ with only a slight steady increase(<1‰). The DO in the tanks remained stable at ~9.5 mg/l through the experiment.

Mesocosm experimentOne mesocosm was extracted at pre-established time intervals: two months (T1) and threemonths (T2) from each tank, namely control (c) and two selected concentrations of Hg-pollut-ants (1 ppm and 100 ppm). The 120-cm3 sediments from each mesocosm were wet-sieved withASW through a 125-μm screen. The>125-μm fraction was the source material from which liv-ing specimens of A. parkinsoniana were picked. The presence of reticulopodial activity or, ifabsent (T2-100 ppm), the presence of cytoplasm in any but the last chamber was the criterionfor specimen selection. The presence of cytoplasm in specimens of the T2-100 ppm group wasnot a sufficient condition for determining specimen viability so further analyses (ultrastruc-tural investigations and confocal-cell imaging) were performed to evaluate cell fitness.

Microscopic analysesTransmission Electron Microscopy. Selected specimens were fixed with 2.5% glutaralde-

hyde (TAAB, England, UK) in ASW for 3 h at 4°C. Foraminiferal tests were then decalcifiedwith 0.1 M EDTA for 36 h. After 5 washings with ASW, foraminiferal specimens were post-fixed with 1% osmium tetroxide (OsO4; EMS, Hatfield, PA) in ASW for 2 h at room tempera-ture (RT). Following 5 washings, specimens were dehydrated in a graded series of ethanolbaths, from 50% to 100% and immersed in propylene oxide (EMS, Hatfield, PA) twice, each for10 minutes. Subsequently, they were embedded in epoxy resin by using increasing concentra-tion of resin (Durcupan Araldite, SIGMA, UK). Foraminifera were ultimately sectioned usingan ultramicrotome (LKB, 2088 UltrotomeV). Thick sections of 1 μm were stained with 1%toluidine blue in distilled water at 60°C to provide light-microscope-level overviews of wholesections. Thin sections (100 nm), collected on 300-mesh nickel grids, were stained with uranylacetate and lead citrate and finally observed with a Philips CM10 transmission electron

Mercury-Pollution and Ammonia parkinsoniana

PLOSONE | DOI:10.1371/journal.pone.0162401 September 7, 2016 3 / 14

microscope at 80 KV [20]. Lipid was quantified by calculating the area of 200 lipid vesicles, onaverage, distributed in the last whorl excluding the final two chambers (last two formed; i.e.,youngest two) from 3–6 selected specimens (200–350 μm in diameter) for the two Hg concen-trations, plus controls, at both time points. The Mann-Whitney U test, a nonparametric test,was used to check for significant differences between mean lipidic dimension among experi-mental conditions (concentration and time).

Confocal and light microscopy. Selected specimens of A. parkinsoniana were incubatedwith Nile Red (NR) or Acridine Orange (AO) at T1 and T2 and then analyzed with CLSM.

Nile Red is a phenoxazine dye used on living and fixed cells to localize and quantify neutraland polar lipids [21–22]. The absorption and fluorescence properties of NR are known to besensitive to environmental factors such as polarity. Polar lipids (i.e., phospholipids), which aremostly present in membranes, fluoresce red (emission> 590 nm) whereas neutral lipids (ester-ified cholesterol and triglycerides), which are present in LD, fluoresce yellow (570–590 nm)[23–24]. NR was used on specimens of A. parkinsoniana to detect membranous vesicles and onwhole specimens to compare the lipidic distributions of untreated and Hg-treated conditionsby confocal microscopy. For NR microscopy, A. parkinsoniana specimens were fixed in 2%paraformaldehyde for 2 hours, then washed in ASW and decalcified with EDTA (0.1 M) for 48hours to remove the foraminiferal test. Following decalcification, specimens were rinsed inASW, transferred to MatTek glass bottom chambers (MatTek Corporation, Ashland, MA) andNR was added at the final concentration of 3 μg/ml for 40 min at RT. Using CLSM, specimenswere subject to blue excitation (488 nm) and analyzed separately for yellow and red emissions.Quantitative analyses of NRMean Fluorescence Intensity (MFI) were performed using ImageJsoftware (NIH, Bethesda, MD), imaging the specimens at the same magnification, and deter-mining the MFI of all the selected pixels. Subsequently, yellow MFI values were converted toarbitrary units (A.U.) setting the first image of the first control specimen at 100. The fluores-cence emission of LD was calculated in the yellow channel of 3–5 selected specimens (200–350 μm in diameter) for each concentration.

The pH-sensitive dye acridine orange (AO) was used to detect and quantify acidic vesicularorganelles in A. parkinsoniana specimens by confocal microscopy [25–26]. AO is a cell-perme-able fluorescent dye that labels DNA and cytoplasm bright green whereas RNA and acidic vac-uoles appear red. It can also enter acidic compartments and organelles, such as lysosomes andautolysosomes, where it becomes protonated and sequestered [25]. When AO is bound to acidcompartments, such as lysosomes and acidic vacuoles, it emits red fluorescence (>650 nm)with intensity proportional to the acidity degree. For confocal live imaging, A. parkinsonianaspecimens were transferred to MatTek glass bottom chambers (MatTek corporation, Ashland,MA) and then stained with AO (150 ng/ml) and incubated for 40 minutes at RT. Green andred fluorescence emissions illuminated with blue (488 nm) excitation light were analyzed byconfocal microscopy. Because AO must be used on living specimens and A. parkinsonianatests may interfere with dye uptake, only peripheral signals were observed (cytoplasm in greenand acidic vesicles in red). Quantitative analyses of AO red MFI were performed on 3–5 speci-mens (200–350 μm in diameter) for control and 100 ppm treatments. Image analyses were car-ried out by determining the red MFI of all the selected pixels of the imaged specimens.Subsequently, the red MFI was converted to arbitrary units (A.U.) setting the first image of thefirst control specimen at 100. Epifluorescence and bright field (BF) microscopies were per-formed using a CLSM (Leica TCS SP5 II confocal microscope, Leica Microsystems) with 488-,543- and 633-nm illumination and oil-immersion objectives.

ESEM-SEM and EDS. Embedded specimens used for TEM were observed, as a whole,with an environmental scanning electron microscope (FEI ESEM, Quanta 200) to qualitativelycharacterize the presence of Hg. The ESEM, coupled with energy dispersive X-ray spectrometry

Mercury-Pollution and Ammonia parkinsoniana

PLOSONE | DOI:10.1371/journal.pone.0162401 September 7, 2016 4 / 14

(EDS), was used to assess the elemental composition of particles in the cytoplasm. The EDS is atechnique employed to collect and determine the energy and the number of X-rays that aregiven off by atoms in a material [27–28]. Observations were conducted in low vacuum (0.2–1.2Torr) at 10-mm working distance using secondary and backscattered electron modes withenergy varying from 12 to 25 kV. A live counting time of 100 seconds, with spots’mode from 3to 5, was used for elemental mapping.

Additionally, thin sections (~100 nm) were mounted onto freshly peeled mica adhered to anSEM stub with double stick tape. The sections were then coated with about 10 nm of carbon usinga Cressington 208 carbon coater. The sample was then imaged and elemental maps were obtainedusing a Hitachi S4800 FEG-SEM at 20 keV equipped with an EDS at the Oak Ridge National Labo-ratory. The upper secondary detector was used for image acquisition at a working distance of 7–8mm. EDS maps were obtained using 64 sweeps with a 200 ms dwell time per pixel at a resolutionof 256×200 pixels. Maps were plotted using the net intensities of the selected regions of interestwith the nearest 3 × 3 pixels averaged to improve signal to noise in the element map images.

Results

Lipid characterizationImages obtained visa CLSM showed enhanced accumulation of cellular lipids in Hg-treated A.parkinsoniana at T1 and T2, compared to control specimens (Fig 1A–1D). In untreated (con-trol; no Hg exposure) A. parkinsoniana specimens, a compact and uniform lipid distributionwas found (Fig 1A) whereas, at increasing Hg concentration at both T1 and T2, enhanced yel-low fluorescence, in particular for the 100-ppm Hg treatment, was observed (Fig 1B–1D). NRquantification by ImageJ software confirmed this observation (Fig 1E), where NR yellow MFImarkedly increased in Hg-treated specimens both in a time- and concentration-dependentmanner. Negligible changes were observed between T1 and T2 among treatments (controls,1 ppm, and 100 ppm). On the other hand, differences were evident between control and treat-ment, namely 1 ppm and 100 ppm at both T1 and T2 (Fig 1E).

Diameters from a total of 1295 LD from 27 specimens were measured from TEM images(S1 Appendix). On the basis of the Mann-Whitney U test, no significant differences in diame-ter were found among all the six conditions except for T2-100 ppm, which exhibited signifi-cantly larger LD (S2 Appendix).

Lysosomal compartment characterizationIn order to evaluate the acidic compartments (lysosomes, autolysosomes and other acidic vesicu-lar organelles) of A. parkinsoniana in response to Hg exposure, several specimens from eachexperimental condition were probed with AO (Fig 2A–2D). At T1, specimens exposed to100-ppm Hg displayed more conspicuous acidic compartments, essentially due to vesicle volume(dimension and, as a consequence, total fluorescence intensity), when compared to those of con-trol specimens (Fig 2E and 2F). Some 100 ppm-exposed specimens also exhibited diffuse redfluorescence in their chambers, suggesting cytoplasmic acidification (Fig 2C). These specimenswere excluded from AOMFI quantitative analyses (Fig 2F). Mercury-exposed specimens at T2appeared to show an acidic compartment decrease as marked by reduction of red vacuoles sug-gesting a loss of lysosomal activity, a possible signal of recent or imminent death (Fig 2D and 2F).

Localization and distribution of HgOn the basis of ESEM-EDS observations, it was possible to observe the presence and distribu-tion of Hg in different parts of the A. parkinsoniana (Fig 3). As expected, no Hg was found in

Mercury-Pollution and Ammonia parkinsoniana

PLOSONE | DOI:10.1371/journal.pone.0162401 September 7, 2016 5 / 14

control specimens. In Hg-incubated specimens, Hg was commonly found in all chambers (notshown) in the basal part of pores (Fig 3C and 3D) and in the foramen (Fig 3E), occurring as adiffuse signal corresponding to the organic linings. Mercury also occurred in the form of gran-ules (1–2 μm) in the cytoplasm, mainly in the youngest chambers (Fig 3A and 3B). The ele-mental analysis revealed the presence of Hg in first and secondary peaks (Fig 3F). The brightlarge vacuoles (Fig 3A and 3B) in the cytoplasm appear to have a lipid nature as confirmed byEDS spectrum revealing the main presence of osmium. The presence of osmium was an artifactresulting from the fixation protocol for optical and electron microscopy, where it is commonlyused to fix and to stain lipids.

The SEM-EDS element maps on thin sections confirm the presence of localized Hg in thebasal part of the pores (S2 Fig). The Hg-rich region was observed as a lighter strip in the sec-ondary emission (SE) SEM image (S2A and S2B Fig). The lighter contrast is a result of back-scattered electrons from the heavier (higher atomic number) Hg atoms detected by the upperpolepiece detector (S2C Fig). The element maps reveal the localized nature of the Hg on theouter lining just below the decalcified test region (S2D Fig). This is supported by the carbondepletion and calcium presence just to the left of the Hg strip. It is possible that the localizedHg could have been a HgS complex. However, since the EDS peak positions of the Hg M andthe S K peaks overlap, it is difficult to exactly characterize the elemental composition. It is cer-tain that this region contains significant amounts of Hg due to the brighter contrast comparedto the background in the SE image (S2A–S2D Fig). Further studies are required to resolve theseunknowns.

Fig 1. Effect of Hg exposure on lipid distribution of A. parkinsoniana labeled with NR. Epifluoresecence micrographs of single opticalsections showing overlay of NR yellow and red fluorescence for (A) T1- control, (B) T1-1 ppm, (C) T1-100 ppm and (D) T2-100 ppm, Bars: 20 μm.(E) Histogram of yellow Mean Fluorescence Intensity (MFI) expressed in arbitrary units (A.U.) for the three treatments over time (control, 1 ppm,and 100 ppm at both T1 and T2). Error bars indicate ± standard error of the mean.

doi:10.1371/journal.pone.0162401.g001

Mercury-Pollution and Ammonia parkinsoniana

PLOSONE | DOI:10.1371/journal.pone.0162401 September 7, 2016 6 / 14

DiscussionHeavy metals occur in ecosystems from both natural sources and human activities, with largevariations in concentrations [29]. Some heavy metals are essential to life, but they can becometoxic through bioaccumulation and biomagnification [30]. Membrane transport or passive diffu-sion of neutral chemical species are reported as pathways to incorporate heavy metals in cells[31]. Heavy metals might penetrate foraminiferal cell membranes together with food and/or maybe incorporated from seawater by membrane transport [32]. Among heavy metals, Hg is consid-ered to be highly toxic for living organisms [30]. Mercury is not biodegradable but bioaccumu-lates in the food web [33]. In particular, uncharged Hg species, including mercury sulfide (HgS),dimethylmercury ((CH3)2Hg), and inorganic mercury (HgCl2), are shown to cross cell mem-branes [34]. Methylmercury has been reported as the most toxic species of Hg because it can eas-ily cross biological membranes; though inorganic Hg, which is capable of passing biologicalbarriers with difficulty, is also toxic but at concentrations higher than methylmercury [35]. Theconcentration and form of Hg in aquatic organisms are controlled by passive uptake of lipophilicchloride complexes (both CH3HgCl and HgCl2) across the cell’s lipid bilayer in phytoplankton[36], whereas an active uptake of Hg(II) was suggested for iron-reducing bacteria (FeRB)Geobac-ter sulfurreducens and sulfate-reducing bacteria (SRB) Desulfovibrio desulfuricans (strain ND132)[37]. More importantly, in phytoplankton, Hg(II) is mainly bound to membranes whereas meth-ylmercury is mostly accumulated inside the cell, in the cytoplasm [36, 38].

Fig 2. Effect of Hg exposure on acidic compartmentalization of A. parkinsoniana labeled with AO. Epifluorescence micrographs of single opticalsections showing overlay of AO green and red fluorescence for (A) T1-control, (B,C) T1-100 ppm, and (D) T2-100 ppm, Bars: 20μm. (E) Histogram ofmaximum dimension (diameter) of acidic vesicles. (F) Histogram of red Mean Fluorescence Intensity (MFI) expressed in arbitrary units (A.U.) forcontrol and 100 ppm at both T1 and T2. Error bars indicate ± standard error of the mean.

doi:10.1371/journal.pone.0162401.g002

Mercury-Pollution and Ammonia parkinsoniana

PLOSONE | DOI:10.1371/journal.pone.0162401 September 7, 2016 7 / 14

On the basis of different methodology, the concentrations of Hg used in our experimentwere higher than those considered by [13, 14] that directly exposed the foraminifer Rosalinaleei in culture and reported specimens living in concentrations up to 260 ng/l (260 ppt). Con-trary to these culture experiments, the present study used higher Hg concentrations and alsoensured that the Hg-ASWmixtures permeated the sediments to expose the foraminifera livingtherein. Although different culture experiments have been conducted to document the effect ofheavy metal pollutants on foraminifera (i.e., [10, 13, 14]), to our knowledge, no study has ever

Fig 3. Micrographs showing presence of Hg in A. parkinsoniana specimens (T2-100 ppm). (A) young chamber containingvacuoles; (B) high magnification of a young chamber; (C,D) basal part of pores; (E) foramen/septum. (F) Example EDSspectrum taken with a spot size of 4 on cross (E). Arrows mark the occurrence of Hg. Scale bar: (A) 10 μm; (B) 1.5 μm; (C)2.5 μm; (D) 0.8 μm; (E) 5 μm.

doi:10.1371/journal.pone.0162401.g003

Mercury-Pollution and Ammonia parkinsoniana

PLOSONE | DOI:10.1371/journal.pone.0162401 September 7, 2016 8 / 14

revealed the direct presence of metals within the cytoplasm of foraminifera. The occurrence ofHg within the A. parkinsoniana cytoplasm suggests that the pollutant crossed the foraminiferalcell membrane in some manner. In particular, Hg was found in all foraminiferal chambers,from the oldest to the youngest and has been mainly found as small particles dispersed in thecytoplasm, and more importantly diffuse at the basal part of pores and along the organiclining.

The effects of pollutants on foraminiferal ultrastructure are not new to science. Indeed,numerical increases of LD characterized by a more electron-dense core, proliferation of resid-ual bodies, fibrillar and large lipidic vesicles, a thickening of the organic lining, mitochondrialdegeneration, and autophagosome proliferation are among the most common documentedcytological modifications (i.e., [9–11]). Unfortunately, these studies failed to document theoccurrence of pollutant within the cytoplasm and the documented effects remain circumstan-tial and speculative. The present paper, instead, documents the presence and potential distribu-tion of Hg within A. parkinsoniana cytoplasm and its possible cytological alterations.

Nile Red distinguished neutral lipids from polar ones (Fig 1A–1D) [39]. In our highest Hg-concentration, we found that more accumulation of neutral lipids occurred in the form of LDcompared to lower Hg or controls. Lipid droplets have fundamental roles in a cell’s metabolismand are commonly related to the cell’s requirement to store excess lipid in neutral-lipid oils liketriacylglycerols and sterol esters [40]. Increasing numbers of LD have been reported in lichenspecies (Evernia prunastri and Xanthoria parietina) as an effect of dust pollution from acement industry [41] and in rat hepatocytes in response of Bisphenol A exposure [42]. Anincrease in the number of LD coupled with their dimension in the hepatocytes of grey mullet(Mugil cephalus) was related to heavy-metal pollution [43]. Similar results were documented inthe hepatocytes of silver catfish (Rhamdia quelen) exposed to sublethal lead concentration[44], in Daphnia similis as a response to nanowire exposure [45] and in male rockfish (Sebastis-cus marmoratus) after paclobutrazol exposure [46]. Decreasing numbers and dimensions ofLD in the liver of fish (male guppies Poecilia reticulata) were associated with the addition ofcysteine in their diet to reduce Hg concentration [47]. Increased number and size of lipid vesi-cles were detected in the liver of HgCl2-treated zebrafish (Danio rerio) and confirmed by thetranscriptome analysis on the regulation of fatty acid synthesis and of mitochondrial fatty acidbeta-oxidation [48]. Lipid droplets are also hypothesized to sequester toxicants in order to pro-tect cells [49–50].

Our results suggest an increased number of LD associated with exposure to high Hg concen-trations (100 ppm). This observation is supported by increased values of yellow fluorescencemarking neutral lipid occurrence associated with higher concentrations of Hg. There was nostatistically significant increase in the dimensions of lipids with respect to Hg except for signifi-cantly larger lipids in the T2-100 ppm treatment. It might be speculated that Hg-pollutionmight promote the proliferation of higher numbers of neutral lipids but not affect their dimen-sion. Although based on a qualitative approach, [10] reported increased numbers and dimen-sions of lipid in two foraminiferal species (Ammonia beccarii and Ammonia tepida) due to Cucontamination. The exact implications and consequences of increases in LD numbers have notbeen explored in our present study.

All control specimens treated with AO were characterized by a weak, diffuse, green fluores-cence of cytoplasm with some dispersed, low numbers and reduced dimension of red vesicles(lysosomes). Lysosomes, acidic membrane-bound organelles containing hydrolytic enzymes,are devoted to digestion of biomolecules, redundant or damaged organelles and proteins aspart of autophagic cellular turnover [51–52]. A wide array of chemicals including metal ions(Fe, Cu and Hg), PAHs, nanoparticles can be effectively sequestered and accumulated by

Mercury-Pollution and Ammonia parkinsoniana

PLOSONE | DOI:10.1371/journal.pone.0162401 September 7, 2016 9 / 14

lysosomes [53] that might also represent the most important site of metal detoxification ineukaryotes [52].

Our AO results document that increased concentrations of Hg initially promote the prolif-eration of lysosomes both in terms of number and dimension. Increased production of lyso-somes and lysosomal accumulation of Hg have been described in cells of insects (Aedesalbopictus) and slugs (Arion ater) [54–55]. Lysosomal proliferation as a detoxification mecha-nism induced by metal contamination has been reported in mussels, oysters, marine dinoflagel-late and foraminifera (i.e., [10, 56] and reference therein). More specifically, inorganic Hg afterhaving formed a complex with selenium or cysteine was reported to accumulate in lysosomesin the liver of the frog Rana ridibunda [57]. The observed lysosomal compartment enlargementis commonly associated with the formation of secondary lysosomes (such as autolysosomes)and degradation process intensification, and may be related to Hg-dependent cell damage [58].Curiously, at the highest Hg concentrations in our study, lysosomal membrane destabilizationlikely induced the acidification of chambers early in the experiment (T1-100 ppm; Fig 2C)whereas the very low levels of lysosomal activity manifest as AOMFI that occurred at the endof the experiment (T2-100 ppm) may have been due to specimen death. Pollutants are knownto destabilize the membranes of lysosomes causing the release of hydrolytic enzymes [59], asreported in eukaryotes [60] including protists [61].

ConclusionThis study examined the effect and bioaccumulation of Hg on a benthic foraminiferal species.Our experiment coupled with different microscopy techniques revealed, for the first time, thepresence of Hg in the cytoplasm of A. parkinsoniana after exposure to elevated Hg in seawater.Our observations show that: (i) an increase in the number of LD, which have been reported asa detoxifying organelle where pollutants might be sequestered and biologically inactivated, inresponse to Hg exposure; (ii) Hg appears to promote the proliferation of lysosomes both interms of number and dimension up to membrane destabilization when the lysosomal contentslikely were released into the cytosol ultimately leading to cell death; (iii) A. parkinsonianaappeared to accumulate Hg in the organic lining at the basal part of pores and in the foramen/septa as well as in cytoplasm, mainly in the younger chambers.

Supporting InformationS1 Appendix. Lipid dimension counting.(XLSX)

S2 Appendix. Mann-Whitney U Test for differences lipid dimension. Significant differencesare marked in bold.(XLS)

S1 Fig. Typical paired NR confocal images. (A) Bright Field image. (B) Neutral lipids (triglyc-erides, esters of cholesterol and free fatty acids) in yellow. (C) Polar lipids (phospholipids,sphingolipids and non-esterified cholesterol) in red. (D) Merged Bright Field, yellow and redchannels.(TIF)

S2 Fig. (A) Secondary emission image of a microtomed sample, organic lining (arrows),deposited on freshly peeled mica and coated with 10 nm of carbon to prevent sample charging.Included is a net elemental cross-section with a 10×10 pixel average of the sample from EDSmapping at 20 kV. (B) Elemental cross-sections showing the increase of Hg in the brighterregion of the SEM image. The presence of Al, Si, K, and O comes primarily from the underlying

Mercury-Pollution and Ammonia parkinsoniana

PLOSONE | DOI:10.1371/journal.pone.0162401 September 7, 2016 10 / 14

mica (muscovite KAl2(AlSi3)O10(OH)2). (C) EDS spectra from the entire SEM image region(shown in A) showing the presence of Hg. (D) EDS maps of the SEM image region showingthe depletion of carbon and the localized presence of Hg. The EDS maps are taken from the netintensities of the element regions. Each point is the average of 3×3 pixels.(TIF)

AcknowledgmentsThe authors are grateful to the Academic Editor of PLOS ONE Yi Hu and two anonymousreviewers for their thoughtful and valuable comments that have greatly improved the paper.We warmly thank Shawn Reeves for SEM training, Carla Bucci for technical assistance andGianluca De Grandis and Fabio Principi of the Agenzia Per La Protezione Ambientale DelleMarche (ARPAM) for sampling assistance.

Author Contributions

Conceptualization: FF RC.

Formal analysis: FF NP RDM.

Investigation: FF DC EC BC FMG BG JRE AMJ EMP LZ.

Methodology: FF.

Resources: PG SP RC.

Visualization: FF DC EC JMB BG.

Writing – original draft: FF DC EC JMB BG.

Writing – review & editing: FF DC EC JMB BG.

References1. Goldstein ST. Foraminifera: a biological overview. In: Sen Gupta BK, editor. Modern foraminifera:

Kluwer Academic Publishers; 1999. pp. 37–55.

2. Sen Gupta BK. Introduction to modern foraminifera. In: Sen Gupta BK, editor. Modern foraminifera:Kluwer Academic Publishers; 1999. pp. 3–6.

3. Lee JJ, Anderson OR. Biology of foraminifera, 1st ed. London: Academic Press; 1991.

4. Frontalini F, Coccioni R. Benthic foraminifera as bioindicators of pollution: A review of Italian researchover the last three decades. Rev Micropaleontol. 2011; 54(2):115–27.

5. Schönfeld J, Alve E, Geslin E, Jorissen F, Korsun S, Spezzaferri S, et al. The FOBIMO (FOraminiferalBIo-MOnitoring) initiative-Towards a standardised protocol for soft-bottom benthic foraminiferal moni-toring studies. Mar Micropaleontol. 2012; 94–95:1–13.

6. Alve E. Benthic foraminiferal responses to estuarine pollution: a review. J Foraminiferal Res. 1995; 25(3):190–203.

7. Yanko V, Arnold AJ, Parker WC. Effects of marine pollution on benthic Foraminifera. In: Sen Gupta BKeditor. Modern Foraminifera. Dordrecht: Kluwer Academic Publishers; 1999. pp. 217–35.

8. Bresler V, Yanko V. Chemical ecology: a new approach to the study of living benthic epiphytic foraminif-era. J Foraminiferal Res. 1995; 25(3):267–79.

9. Morvan J, Le Cadre V, Jorissen F, Debenay JP. Foraminifera as potential bio-indicators of the "Erika"oil spill in the Bay of Bourgneuf: Field and experimental studies. Aquat Living Resour. 2004; 17(3SPEC. ISS.):317–22.

10. Le Cadre V, Debenay JP. Morphological and cytological responses of Ammonia (foraminifera) to cop-per contamination: Implication for the use of foraminifera as bioindicators of pollution. Environ Pollut.2006; 143(2):304–17. PMID: 16442682

Mercury-Pollution and Ammonia parkinsoniana

PLOSONE | DOI:10.1371/journal.pone.0162401 September 7, 2016 11 / 14

11. Frontalini F, Curzi D, Giordano FM, Bernhard JM, Falcieri E, Coccioni R. Effects of lead pollution onAmmonia parkinsoniana (foraminifera): Ultrastructural and microanalytical approaches. Eur J Histo-chem. 2015; 59(1). doi: 10.4081/ejh.2015.2460

12. Anderson OR, Lee JJ. Cytology and fine structure. In: Lee JJ, Anderson OR, editors. Biology of Forami-nifera. Academic Press: London; 1991. pp. 335–356.

13. Saraswat R, Kurtarkar SR, Mazumder A, Nigam R. Foraminifers as indicators of marine pollution: A cul-ture experiment with Rosalina leei. Mar Pollut Bull. 2004; 48(1–2):91–6. PMID: 14725879

14. Nigam R, Linshy VN, Kurtarkar SR, Saraswat R. Effects of sudden stress due to heavy metal mercuryon benthic foraminifer Rosalina leei: Laboratory culture experiment. Mar Pollut Bull. 2009; 59(8–12):362–8. doi: 10.1016/j.marpolbul.2009.08.014 PMID: 19748104

15. Clarkson TW, Magos L. The toxicology of mercury and its chemical compounds. Crit Rev Toxicol.2006; 36(8):609–62. PMID: 16973445

16. Eisler R. Mercury Hazards to Living Organisms. New York: CRC Press; 2006.

17. Gu B, Bian Y, Miller CL, DongW, Jiang X, Liang L. Mercury reduction and complexation by naturalorganic matter in anoxic environments. Proc Natl Acad Sci U S A. 2011; 108(4):1479–83. doi: 10.1073/pnas.1008747108 PMID: 21220311

18. Frontalini F, Coccioni R. Benthic foraminifera for heavy metal pollution monitoring: A case study fromthe central Adriatic Sea coast of Italy. Estuar Coast Shelf Sci. 2008; 76(2):404–17.

19. Ciacci C, Canonico B, BilanicovaD, Fabbri R, Cortese K, Gallo G, et al. Immunomodulation by differenttypes of N-oxides in the hemocytes of the marine bivalveMytilus galloprovincialis. PLoS ONE. 2012; 7(5). doi: 10.1371/journal.pone.0036937

20. Salucci S, Battistelli M, Burattini S, Squillace C, Canonico B, Gobbi P, Papa S, Falcieri E. C2C12 myo-blast sensitivity to different apoptotic chemical triggers. Micron. 201; 41(8):966–73. doi: 10.1016/j.micron.2010.07.002 PMID: 20674376

21. Greenspan P, Fowler SD Spectrofluorometric studies of the lipid probe, nile red. J Lipid Res. 1985; 26(7):781–9. PMID: 4031658

22. Sackett DL, Wolff J. Nile red as a polarity-sensitive fluorescent probe of hydrophobic protein surfaces.Anal Biochem. 1987; 167(2):228–34. PMID: 3442318

23. Greenspan P, Mayer EP, Fowler SD. Nile red: a selective fluorescent stain for intracellular lipid droplets.J Cell Biol. 1985; 100(3):965–73. PMID: 3972906

24. Diaz G, Melis M, Batetta B, Angius F, Falchi AM. Hydrophobic characterization of intracellular lipids insitu by Nile Red red/yellow emission ratio. Micron. 2008; 39(7):819–24. doi: 10.1016/j.micron.2008.01.001 PMID: 18329888

25. Traganos F, Darzynkiewicz Z. Lysosomal proton pump activity: supravital cell staining with acridineorange differentiates leukocyte subpopulations. Methods Cell Biol. 1994; 41:185–94. PMID: 7532261

26. Mascotti K, McCullough J, Burger SR. HPC viability measurement: trypan blue versus acridine orangeand propidium iodide. Transfusion. 2000(6: ); 40:693–6. PMID: 10864990

27. Goldstein J, Newbury DE, Joy DC, Lyman CE, Echlin P, Lifshin E, et al. Scanning Electron Microscopyand X-Ray Microanalysis. 3rd ed. New York: Kluwer Academic/Plenum Publishers; 2003.

28. Reed SJB. Electron Microprobe Analysis and Scanning Electron Microscopy in Geology. 2nd ed. NewYork: Cambridge University Press; 2005.

29. Stankovic S, Stankovic RA. Bioindicators of toxic metals. In: Lichtfouse E, Schwarzbauer J, Robert D,editors. Environmental chemistry for a sustainable world. Berlin: Springer; 2013. pp. 151–228.

30. Stankovic S, Kalaba P, Stankovic AR. Biota as toxic metal indicators. Environ Chem Lett. 2014; 12(1):63–84.

31. Delnomdedieu M, Boudou A, Geordescauld D, Dufourc EJ. Specific interactions of mercury chloridewith membranes and other ligands as revealed by mercury-NMR. Chem Biol Interact 1992; 81(3):243–69. PMID: 1540995

32. Yanko V, Ahmad M, Kaminski M. Morphological deformities of benthic foraminiferal tests in response topollution by heavy metals: Implications for pollution monitoring. J Foraminiferal Res. 1998; 28(3):177–200.

33. Raspanti E, Cacciola SO, Gotor C, Romero LC, García I. Implications of cysteine metabolism in theheavy metal response in Trichoderma harzianum and in three Fusarium species. Chemosphere 2009;76(1):48–54. doi: 10.1016/j.chemosphere.2009.02.030 PMID: 19298998

34. Gutknecht J. Inorganic mercury (Hg2+) transport through lipid bilayer membranes. J Membr Biol. 1981;61(1):61–6.

35. Gentès S, Maury-Brachet R, Feng C, Pedrero Z, Tessier E, et al. Specific effects of dietary methylmer-cury and inorganic mercury in zebrafish (Danio rerio) determined by genetic, histological, and

Mercury-Pollution and Ammonia parkinsoniana

PLOSONE | DOI:10.1371/journal.pone.0162401 September 7, 2016 12 / 14

metallothionein responses. Environ Sci Technol. 2015; 49(24), 14560–9. doi: 10.1021/acs.est.5b03586 PMID: 26509634

36. Mason RP, Reinfelder JR, Morel FMM. Uptake, toxicity, and trophic transfer of mercury in a coastal dia-tom. Environ Sci Technol. 1996; 30(6):1835–45.

37. Schaefer JK, Rocks SS, ZhengW, Liang L, Gu B, Morel FM. Active transport, substrate specificity, andmethylation of Hg(II) in anaerobic bacteria. Proc Natl Acad Sci U S A. 2011; 08(21):8714–9.

38. Gu B, Mishra B, Miller C, WangW, Lai B, Brooks SC, et al. X-ray fluorescence mapping of mercury onsuspended mineral particles and diatoms in a contaminated freshwater system. Biogeosciences. 2014;11:5259–67.

39. Diaz G, Melis M, Musin A, Piludu M, Piras M, Falchi AM. Localization of MTT formazan in lipid droplets.An alternative hypothesis about the nature of formazan granules and aggregates. Eur J Histochem.2007; 51(3):213–18. PMID: 17921117

40. Thiam AR, Farese RV Jr, Walther TC. The biophysics and cell biology of lipid droplets. Nat Rev MolCell Biol. 2013; 14(12):775–86. doi: 10.1038/nrm3699 PMID: 24220094

41. Paoli L, Guttová A, Grassi A, Lackovičová A, Senko D, Sorbo S, et al. Ecophysiological and ultrastruc-tural effects of dust pollution in lichens exposed around a cement plant (SW Slovakia). Environ Sci Pol-lut Res Int. 2015; 22(20):15891–902. doi: 10.1007/s11356-015-4807-x PMID: 26044142

42. Grasselli E, Cortese K, Voci A, Vergani L, Fabbri R, Barmo C, et al. Direct effects of Bisphenol A on lipidhomeostasis in rat hepatoma cells. Chemosphere. 2013; 91(8):1123–29. doi: 10.1016/j.chemosphere.2013.01.016 PMID: 23399309

43. Vasanthi LA, Revathi P, Mini J, Munuswamy N. Integrated use of histological and ultrastructural bio-markers inMugil cephalus for assessing heavy metal pollution in Ennore estuary, Chennai. Chemo-sphere. 2013; 91(8):1156–64. doi: 10.1016/j.chemosphere.2013.01.021 PMID: 23415490

44. Muñoz L, Weber P, Dressler V, Baldisserotto B, Vigliano FA. Histopathological biomarkers in juvenilesilver catfish (Rhamdia quelen) exposed to a sublethal lead concentration. Ecotoxicol Environ Saf.2015; 113:241–7. doi: 10.1016/j.ecoenv.2014.11.036 PMID: 25521338

45. Artal MC, Holtz RD, Kummrow F, Alves OL, Umbuzeiro Gde A. The role of silver and vanadium releasein the toxicity of silver vanadate nanowires toward Daphnia similis. Environ Toxicol Chem. 2013; 32(4):908–12. doi: 10.1002/etc.2128 PMID: 23341191

46. Sun L, Li J, Zuo Z, Chen M, Wang C. Chronic exposure to paclobutrazol causes hepatic steatosis inmale rockfish Sebastiscus marmoratus and the mechanism involved. Aquat Toxicol. 2013; 126:148–53. doi: 10.1016/j.aquatox.2012.11.002 PMID: 23202249

47. MokWJ, Hatanaka Y, Seoka M, Itoh T, Tsukamasa Y, Ando M. Effects of additional cysteine in fish dieton mercury concentration. Food Chem. 2014; 147:340–5. doi: 10.1016/j.foodchem.2013.09.157 PMID:24206728

48. Ung CY, Lam SH, Hlaing MM, Winata CL, Korzh S, Mathavan S, et al. Mercury-induced hepatotoxicityin zebrafish: In vivo mechanistic insights from transcriptome analysis, phenotype anchoring and tar-geted gene expression validation. BMCGen. 2010; 11(1):212.

49. Murphy G Jr, Rouse RL, Polk WW, HenkWG, Barker SA, Boudreaux MJ, et al. Combustion-DerivedHydrocarbons Localize to Lipid Droplets in Respiratory Cells. Am J Respir Cell Mol Biol. 2008; 38(5):532–40. PMID: 18079490

50. Rowan-Carroll A, Halappanavar S, Williams A, Somers CM, Yauk CL. Mice exposed in situ to urban airpollution exhibit pulmonary alterations in gene expression in the lipid droplet synthesis pathways. Envi-ron Mol Mutagen. 2013; 54(4):240–9. doi: 10.1002/em.21768 PMID: 23536514

51. Yorimitsu T, Klionsky DJ. Autophagy: molecular machinery for self-eating. Cell Death Differ. 2005; 12(2):1542–1552.

52. Gomiero A, Dagnino A, Nasci C, Viarengo A. The use of protozoa in ecotoxicology: Application of multi-ple endpoint tests of the ciliate E. crassus for the evaluation of sediment quality in coastal marine eco-systems. Sci Total Environ. 2013; 442:534–44. doi: 10.1016/j.scitotenv.2012.10.023 PMID: 23202299

53. Moore MN, Icarus Allen J, McVeigh A. Environmental prognostics: An integrated model supportinglysosomal stress responses as predictive biomarkers of animal health status. Mar Environ Res. 2006a;61(3):278–304.

54. Marigómez I, Soto M, Kortabitarte M. Tissue-level biomarkers and biological effect of mercury on senti-nel slugs, Arion ater. Arch Environ Contam Toxicol. 1996; 31(1):54–62. PMID: 8687990

55. Braeckman B, Raes H. The ultrastructural effect and subcellular localization of mercuric chloride andmethylmercuric chloride in insect cells (Aedes albopictusC6/36). Tissue Cell. 1999; 31(2):223–232.doi: 10.1054/tice.1999.0021 PMID: 18627859

Mercury-Pollution and Ammonia parkinsoniana

PLOSONE | DOI:10.1371/journal.pone.0162401 September 7, 2016 13 / 14

56. Prevot P, Soyer-Gobillard MO. Combined action of cadmium and selenium on two marine dinoflagel-lates in culture, Prorocentrummicans Ehrbg. andCrypthecodinium cohnii Biecheler. J Protozool.1986; 33(1):42–7.

57. Loumbourdis NS, Danscher G. Autometallographic tracing of mercury in frog liver. Environ Pollut 2004.129(2);299–304. PMID: 14987815

58. Klionsky DJ, Abeliovich H, Agostinis P, Agrawal DK, Aliev G, et al. Review: Guidelines for the use andinterpretation of assays for monitoring autophagy in higher eukaryotes. Autophagy. 2008; 4(2):151–75.PMID: 18188003

59. Nicholson S. Lysosomal membrane stability, phagocytosis and tolerance to emersion in the musselPerna viridis (Bivalvia: Mytilidae) following exposure to acute, sublethal, copper. Chemosphere. 2003;52(7):1147–51. PMID: 12820995

60. Moore MN, Allen JI, McVeigh A, Shaw J. Lysosomal and autophagic reactions as diagnostic and pre-dictive indicators of environmental pollutant toxicity in aquatic animals. Autophagy. 2006b; 2(3):217–20.

61. Dondero F, Jonsson H, Rebelo M, Pesce G, Berti E, Pons G, et al. Cellular responses to environmentalcontaminants in amoebic cells of the slime mould Dictyostelium discoideum. Comp Biochem Physiol CToxicol Pharmacol. 2006; 143(2):150–7. PMID: 16513429

Mercury-Pollution and Ammonia parkinsoniana

PLOSONE | DOI:10.1371/journal.pone.0162401 September 7, 2016 14 / 14